Using Conditional MLE for Logistic Regression Models by Lia Lee '17
Wednesday, February 15 at 1:10pm to 1:35pm
Abstract: In inferential statistics, maximum likelihood estimation (MLE) is a common method for estimating the unknown parameters of a statistical model. For logistic regression models, MLE often breaks down when regularity conditions are not met (e.g. when n increases, so does the number of parameters). In such cases, we turn to a conditional likelihood approach where nuisance parameters can be eliminated by conditioning on their sufficient statistics. In this talk, we will introduce the theory behind conditional MLE as well as applications of modeling for binary matched-pairs data in biomedicine.
No recent activity